Ampeg B-18 issue/debug

[aborgman]

I've been using this B-18 head into a LDS 15/6 for the past 4 months or so. It's been performing well, but just in the last couple shows I've noticed it seemed like it was starting to get dirty before it got loud enough.

I thought it might be the room we were playing, or my guitar players new amp drowning me out - but I thought I'd haul it into work and stick it on the 'scope and see how it looked.

Up to ~20 watts - it looks just fine. The signal across the dummy load look just great.

Above that, things start to get ugly. The top half of the wave looks fine all the way up to the point it starts flat topping - but the bottom half has some odd noise burst/oscillation thing going on that gets worse with increasing power levels above ~20 watts.

Just above 20W:


A bit higher:


Just before flat topping (~50W):


...and a close up of the oscillation - it's at ~95kHz.


I pulled all the tubes and cleaned/tightened the sockets.

I swapped power tubes - the burst didn't move.

So it looks like it isn't a bad power tube.



Any thoughts/suggestions before I get around to probing everything?

[BassmanPaul]

You've got instability in there. Have a look at the B+ lines and see if you don't have a failing cap. Trace from the input jack with the scope and see where this problem starts. I don't have the schematic at hand at the moment.

[aborgman]

Originally Posted by BassmanPaul

You've got instability in there.

Yeah, of some odd sort.

Originally Posted by BassmanPaul

Have a look at the B+ lines and see if you don't have a failing cap.

Seems like if it were a B+ issue it would affect both halves of the wave - what mechanism would cause one side to show instability like that?

Originally Posted by BassmanPaul

Trace from the input jack with the scope and see where this problem starts. I don't have the schematic at hand at the moment.

It's fine through the entirety of the preamp section - so it's either the phase splitter or the power section where it's having an issue.

[beans-on-toast]

The first thing that I would try would be to change the phase inverter tube if you have a spare.

Next I would move the wires in the power amp stage with a chopstick while performing the test to see if the oscillation changes in any way.

[aborgman]

Originally Posted by beans-on-toast

The first thing that I would try would be to change the phase inverter tube if you have a spare.

Yeah, that would certainly be my next move if I had one around. Sadly, I don't have a spare - and they're $40+ if you can find one, or you have to drop $20 for a 6GH8 adapter and still buy a tube.

[okcrum]

It's the power amp/OT. You should check the routing of the under chassis wiring. You should clean up any ground connections that are bolted to the chassis as well. If they are riveted, you may have to either solder them to the chassis, bond them with wire or braid, or drill them out and put in screws and lockwashers. I had this problem with the last Leslie PA I restored. Changes in the circulating ground currents can cause chirping like this. Oxidation of the hardware can change the ground currents.

Bad routing of wires or bad grounds can cause positive feedback in this type of amp.

[beans-on-toast]

Well oscillation and lead dress go hand in hand so try chopsticking. Just be carefull not no break any of those old wires in the process.

They have 7199's here for $25.

[aborgman]

Originally Posted by beans-on-toast

Well oscillation and lead dress go hand in hand so try chopsticking. Just be carefull not no break any of those old wires in the process.

Yeah - I poked at every wire surrounding the phase inverter, power tubes, and OT wiring - and it didn't change the output one bit sadly.

I think it's definitely component and not wire related.

Originally Posted by beans-on-toast

They have 7199's here for $25.

Good price! The $7 minimum shipping charge kinda stings though.

[aborgman]

The output of the phase inverter/grids of the power tube are showing the oscillation - and I'm pretty certain it's nice and clean through the entirety of the preamp (need to check again when I get back to work)...

So that looks like it's something in the 7199 or it's related circuitry.

[okcrum]

Originally Posted by aborgman

The output of the phase inverter/grids of the power tube are showing the oscillation - and I'm pretty certain it's nice and clean through the entirety of the preamp (need to check again when I get back to work)...

So that looks like it's something in the 7199 or it's related circuitry.

Maybe, but... There's a feedback loop here that's unintended, and you might be looking at its input side. If you don't see anything unusual at the phase inverter input pin 9, then it's the power amp oscillating. Is the bias supply clean?

[aborgman]

Originally Posted by okcrum

Maybe, but... There's a feedback loop here that's unintended, and you might be looking at its input side.

Certainly unintended feedback, all oscillations are... but I'm weirded out by the burst only occurring on negative side of the wave.

It shows up at both phase inverter outputs, but of course inverted on one side.

Originally Posted by okcrum

If you don't see anything unusual at the phase inverter input pin 9, then it's the power amp oscillating. Is the bias supply clean?

I'll give bias supply a good check tomorrow, I know it's prone to failure. Any other possibilities for a power amp oscillation that moving wires around has no effect on?

[aborgman]

Originally Posted by bassmeknik

"Seems like if it were a B+ issue it would affect both halves of the wave - what mechanism would cause one side to show instability like that?"


Wrong the power supply is really two halves each powering either the positive half or negative half of the waveform. Look for power supply problems in the negative side only (assuming your waveform is shown in proper phase) such as leaking filter cap

Ummmm....

This is a tube amp. It only has a positive supply, it isn't dual rail like most SS amps.

Both halves of the waveform are driven by the same (positive) supply through the center tap of the output transformer.

[okcrum]

Originally Posted by aborgman

I'll give bias supply a good check tomorrow, I know it's prone to failure. Any other possibilities for a power amp oscillation that moving wires around has no effect on?

Bad grounds can cause this, and even PS filter caps that are not quite up to snuff, since there are no small value (usually ceramic) bypass caps in this PA circuit. No small snubber caps to ground at the 7027 plates, either, so it's all on the filter caps to be low impedance even at 95 kHz.

[beans-on-toast]

A potentially bad solder joint or bad cap are worth checking out. It wouldn't hurt to reflow some of the solder joints.

Another thing to try is to bypass the power supply caps with a 0.1 ceramic disk capacitor. You can use a jumper to clip on the cap and test. A bad cap will pass high frequency hash. The smaller cap in parallel will catch it. This would help you identify a cap with a problem.

Although a number of good possibilities have been put forward, more often than not, it is a bad tube that is causing the oscillation.

[aborgman]

So I've managed to trace it all the way back to the first gain stage - above 20-25W it's starts showing up on the output of the first gain stage feeding the tone stack.

It effects both channels, so it isn't the tube... it's got to be coupling from somewhere or a power issue.

All the power looks good - both voltage wise and noise wise. The first stage out of the rectifier is showing normal 120Hz ripple, but no high frequency noise. All the other stages look wonderful.

Bias voltage is good (-51V) and shows no noise.

Chopsticking around the wiring from the input jack->first gain stage->tone control-> second gain stage (which all happens to run back and forth right under the output transformer) showed no change.

Bypassing the the PS caps with a .01uF cap shows no change in the oscillation.

[beans-on-toast]

Channel 2 doesn't have a grid stopper which can make it prone to picking up RF. I like to add a resistor at pin-4 on the tube socket of these amps. Try removing the channel 2 6SL7 and see if it makes a difference.

I had an oscillation in a B-15 preamp that would manifest itself as a howl at high volumes. The gray shielded wires in the pre were in rough shape. I replaced them and it fixed the problem.
Check the wire and solder joints from the grid, pin-1 at the tube socket, to the 1M volume pot.

[aborgman]

Originally Posted by beans-on-toast

Channel 2 doesn't have a grid stopper which can make it prone to picking up RF. I like to add a resistor at pin-4 on the tube socket of these amps. Try removing the channel 2 6SL7 and see if it makes a difference.

Even channel one only barely has grid stoppers - the resistors are mounted way out at the jack, not at the socket.

Originally Posted by beans-on-toast

I had an oscillation in a B-15 preamp that would manifest itself as a howl at high volumes. The gray shielded wires in the pre were in rough shape. I replaced them and it fixed the problem.
Check the wire and solder joints from the grid, pin-1 at the tube socket, to the 1M volume pot.

Mine has the pre-amp section that matches this:
http://www.drtube.com/schematics/ampeg/b12n-jp.gif

With the power amp section that matches this:
http://www.drtube.com/schematics/ampeg/b18n-jp.gif

I am getting tempted to just throw grid stoppers on every grid that doesn't have one...

[okcrum]

Originally Posted by aborgman

I am getting tempted to just throw grid stoppers on every grid that doesn't have one...

Well, that might well fix the symptom for now. Has this amp ever been recapped? It might be time, then start again from there.

Still, this just reeks of bad ground connection to me. GL, whatever you decide to do.

[beans-on-toast]

Originally Posted by aborgman

Even channel one only barely has grid stoppers - the resistors are mounted way out at the jack, not at the socket.

I suppose that it is easier for them to mount then on the input jacks. I was referring to channel 2 when I say that I add a grid resistor at the socket. Better because the wire from the input won't act as antenna that will affect the amp. Not better because the resistor doesn't have a mounting point to the wire-resistor connection just hangs there. Some heat shrink tubing helps.




This is an example of a bad solder joint on a pot tab that is common to see on these amps. Note that the tab is bridged to the pot case with solder. No wire and not in direct contact. Solder is not that good a conductor.




Here is a bad copy of the original for your schematic. There are a few variations. I was referring to a different version (the second schematic) when I made my comment above.

[beans-on-toast]

Originally Posted by okcrum

Well, that might well fix the symptom for now. Has this amp ever been recapped? It might be time, then start again from there.

Still, this just reeks of bad ground connection to me. GL, whatever you decide to do.

Yes, recapping is always a good idea. Especially for a gigging amp if the caps are over 5 years old. It eliminates yet another possibility.

These amps only have one ground point, located at the input jack. All grounds tie to a buss that terminates there. That's why I was saying to check the solder joints. It amounts to the same thing as far as grounds are concerned.